In recent surveys of VA outpatients, prevalence of alcoholism ranged from 13.6% to 31% depending on the defining criteria and group of outpatients studied. Similar surveys of VA inpatients indicate even higher prevalence of 24% to 43% currently alcoholic. Untreated alcoholism is associated with high rates of health care use and increased health care costs. Despite an over- representation of alcoholics among clinic populations and frequent contact with the medical profession, recent data from the Epidemiologic Catchment Area projects indicate that at most, only 18% of all alcoholics receive treatment for this disorder. A key problem is under-recognition and under-diagnosis by physicians. The proposed research is designed to evaluate: a) the effects of a screening intervention on physicians' rates of recognition and diagnosis of alcoholism (particularly at earlier stages), on rates of alcoholism treatment and referral, and on other clinical decisions; and, c) reasons for physicians' decisions concerning alcoholism. The general design of the study is a randomized clinical trial preceded by a prevalence study. Approximately 1328 primary care outpatients will be screened for alcoholism using the Veterans Alcoholism Screening Test (VAST) and CAGE questions in order to yield approximately 186 alcoholics. All alcoholics will also be administered the Addiction Severity Index to measure various aspects of functioning. On the basis of VAST results, alcoholics will be randomly assigned (1:1) to an experimental group, but not for the control group. The clinical trial will evaluate the effects of providing screening feedback to physicians on alcoholism recognition and diagnosis, rates of treatment and referral, and clinical decisions. A 12 month retrospective and 12 month prospective chart review will be conducted to evaluate the pre-versus post-study behavior of the physicians. At the end of the study period, all physicians will be queried in an in-depth structured interview concerning their reasons for confirming or not confirming as a medical problem the alcoholism screening results, and the reasons for subsequent clinical decisions. Potential benefits of this study include improved rates of recognition and diagnosis of alcoholism, improved rates of alcoholism treatment and referral, decreased rates of health care use and a better understanding of physician decisionmaking concerning alcoholics in primary care.